A Self-Heuristic Ant-Based Method for Path Planning of Unmanned Aerial Vehicle in Complex 3-D Space With Dense U-Type Obstacles

Optimal path planning is required in autonomous navigation and intelligent control of the unmanned aerial vehicle (UAV). However, as a kind of common obstacles in complex three-dimensional (3-D) spaces, U-type obstacles may cause UAV to be confused and even lead to a collision or out of control. Although most of the Ant Colony Optimization (ACO) algorithm can generate proper path, solutions to U-type obstacles based on the specific behaviors of each ant are investigated rarely. Hence, different search strategies are studied and a novel ACO-based method called Self-Heuristic Ant (SHA) is proposed in this paper. The whole space is constructed by grid workspace model firstly, and then a new optimal function for UAV path planning is built. To avoid ACO deadlock state (i.e., ants are trapped in U-type obstacles when there is no optional successor node), two different search strategies are designed for choosing the next path node. In addition, the SHA is utilized to improve the ability of the basic ACO-based method. Specifically, besides pheromone update, a new information communion mechanism is fused to deal with the special areas which contain dense obstacles or many concave blocks. Finally, several experiments are investigated deeply. The results show that the deadlock state can be reduced effectively by the designed two different search strategies of ants. More importantly, compared with the conventional fallback strategy, the average number of retreats and the average running time of ACO can be reduced when SHA is applied.